Facilitating Communication Between Languages and Cultures: A Computerized Interface and Knowledge Base

With the support of the Geneva International Academic Network (GIAN), the Information Systems Interfaces Research Group, the UNDL Foundation and the United Nations Institute for Training and Research have joined their efforts on a project aimed at facilitating multilingual and multicultural communication on the Internet.

The World Wide Web represents an formidable tool for communication and information access. With a simple equipment, it is possible to access innumerable documents about a huge variety of topics, from any place around the world. However, despite the abundance of information, languages very often cause problems. Indeed, although more and more webpages are written in other languages than English, the huge majority of pages represent only about ten languages. Thus, depending on the languages mastered by a user, his or her accessible ressources will be more or less restricted.

Several computer translation systems have been created to bring an answer to the languages problem on the Internet. For example, one can cite the translation services of Google (http://www.google.ch/language_tools) and Altavista (http://babelfish.altavista.com/). Nevertheless these kinds of systems have significant drawbacks. In particular the quality of results is often disappointing (in fact, computer translation systems perform well only in particular situations). These systems also only exist for a limited number of languages combinations.

The UNL (Universal Networking Language) is a United Nations initiative to improve communication between languages and cultures on the Internet. The UNL is an artificial, formal and non-ambiguous language; it is independant of natural languages. The role of the UNL is to act as an intermediary. Unlike computer translation systems, the UNL system will not translate a text directly from a source language to a target language. With the UNL system, texts will be written directly in UNL, with the help of appropriate software tools. As UNL is formal and non-ambiguous, the UNL text can then be transformed automatically in a target natural language.

One of the most importants elements of the UNL system is its knowledge base. This knowledge base contains not only a dictionary of "universal" words that represent concepts but also a set of relations between those concepts. All the words that appear in a UNL text must have been previously defined in the UNL knowledge base. The construction and maintenance of this knowledge base raise several interesting problems, because of its particularities (large size, number of contributors and their geographical dispersion, linguistic and cultural issues).

The goal of our project is to create an infrastructure and tools to support the development and the evolution of the UNL knowledge base. Among the aspects that will be studied, one can mention in particular : "collaborative work: how to co-ordinate multiple participants who are located in different places and work asynchronously ? "multicultural and multilingual issues: how to achieve a "global" knowledge base when participants have different languages and cultures ? "infrastructure: which infrastructure is suited to asynchronous work and still ensures a coherent knowledge base ?

This project includes both theoretical and applied research. As there is no known straightforward solution to this kind of problems, theoretical studies will be carried out before the realisation of prototypes and tools. The project corresponds to three distinct yet closely interrelated parts : 1) design and prototyping of an appropriate storage infrastructure for the knowledge base, 2) design and prototyping of various user interfaces to support different types of tasks and users, 3) establishment of a methodology for the collaborative construction of the knowledge base.

Project team

Dr Gilles Falquet, Department of Information Systems, University of Geneva
Prof. Tarcisio Della Senta, President of the UNDL Foundation
Pascal Renaud, "Information Society" Programme, UNITAR
Claire-Lise Mottaz Jiang, Department of Information Systems, University of Geneva
Dr Jacques Guyot, Département de systèmes d'informations, Université de Genève
M. Saïd Radhouani, Département de systèmes d'informations, Université de Genève
Dr. Rodolpho Pinto Da Luz, UNDL Institute in Brasil
Dr. Hiroshi Uchida, UNL Laboratory, United Nations University, Tokyo
Meiying Zhu , UNL Laboratory, United Nations University, Tokyo
Gabriela Tissiani, visiting PhD Student, Department of Information Systems, University of Geneva

The grant provided by the GIAN for this project totals CHF 240'000.